3 Replies Latest reply on Jul 26, 2017 10:33 AM by Shinichiro Murakami

    Need help finding a calculation to dynamically group a dimension

    Joe Geglio

      I'm working with survey data for students' reported destinations upon graduating from a baccalaureate program. The data is broken down by an assortment of outcomes, sorted by % of total. There is a 'year' filter so the user can view the differences in outcomes by year.


      What I want to do is, for a given year, take all the categories in the field that account for <1% of the total and lump them in with the preexisting 'Other' response. However, the number and type of responses that dip below the 1% threshold vary from year to year.


      I came up with the calculation 'Identify<1%', which recategorizes <1% responses as Other, but it's not good enough as I am unable to sum the 'Other's' into a single 'Other' category (see the Table tab), nor am I able to sort by % of total using this calc as a dimension in the rows shelf.


      Any help would be greatly appreciated!